A Corpus-Based Discourse Analysis of News Media on the Busan Seawater Desalination Plant

This study attempts to clarify a cognitive dissonance in a macro debate on a social issue conducted by news media, proposing a corpus-based opinion analysis methodology. For the purpose, the language use difference between five news media reports on the Busan Seawater Desalination Plant Project were analyzed by applying a TFIDF method. Our proposed corpus-based opinion analysis methodology adopting the Handic-Mecab morpheme analysis and TFIDF method is useful to extract keywords automatically from a large amount of text data. It may contribute to found a corpus-based opinion system and further research. In addition, if there exist a conflict on a public project, the proposed corpus-based opinion analysis can be applied to clarify a cognitive dissonance and find a solution from the large amount of debate data. Comparing to the traditional way of investigation on a conflict structure by a researcher recognition, the proposed methodology is a more objective and contribute to a practical project.